What is AWS Lambda Managed Instances?
AWS Lambda Managed Instances is a Lambda option that runs functions on dedicated, EC2-backed instances. You choose the instance type, such as Graviton4 or memory-optimized variants, and AWS handles the full operation: instance lifecycle, operating system and runtime patching, routing, load balancing, and autoscaling. This way AWS Lambda Managed Instances combines the familiar Lambda programming model with the performance and hardware choice of EC2.
The service addresses two limits of standard Lambda. First, each function can use up to 32 GB of memory and access specialized hardware such as GPUs and high-bandwidth networking, which enables AI/ML and HPC workloads. Second, cold starts are eliminated because capacity scales asynchronously based on CPU utilization. As a result, AWS Lambda Managed Instances suits predictable traffic that needs consistent performance.
Core Features
- Dedicated EC2 instances: Run on instance types of your choice, including Graviton4 and memory-optimized types, with up to 32 GB RAM and 16 vCPUs per function.
- Specialized hardware: Access GPUs as well as high-bandwidth and EFA networking for AI/ML inference and HPC.
- Capacity providers: Define VPC configuration, instance requirements, and scaling policies, and act as the security boundary. All functions on one capacity provider must be mutually trusted.
- Scaling without cold starts: Asynchronous scaling based on CPU utilization. Each execution environment handles multiple invocations concurrently, which benefits IO-heavy applications and requires thread safety.
Typical Use Cases
Memory-intensive applications: Workloads that exceed standard Lambda limits run with up to 32 GB RAM on a single function, without you managing instances yourself.
AI/ML inference: Models benefit from GPU instances and high-bandwidth networking. AWS Lambda Managed Instances provides the specialized hardware while keeping the Lambda model.
Predictable production traffic: Applications with steady load avoid cold starts through utilization-based scaling and use EC2 Savings Plans or Reserved Instances to optimize cost.
Benefits
- Lambda programming model with access to larger and specialized hardware
- No cold starts through asynchronous, utilization-based scaling
- Use of EC2 Savings Plans and Reserved Instances to optimize cost for predictable workloads
- AWS manages patching, routing, load balancing, and autoscaling
Integration with innFactory
As an AWS Reseller, innFactory supports you with the adoption and operation of AWS Lambda Managed Instances: from selecting the right instance types to configuring capacity providers and optimizing cost with Savings Plans.
Typical Use Cases
Frequently Asked Questions
What is AWS Lambda Managed Instances?
Lambda Managed Instances runs Lambda functions on dedicated Amazon EC2 instances of your choice, such as Graviton4 or memory-optimized types. AWS handles instance lifecycle, OS and runtime patching, routing, load balancing, and autoscaling. You keep the Lambda programming model while gaining access to larger and specialized hardware.
When should I use AWS Lambda Managed Instances?
Use it for memory-intensive applications up to 32 GB RAM, AI/ML inference on GPUs, HPC workloads with high-bandwidth networking, and predictable traffic where you want to avoid cold starts. It also suits existing Lambda functions that need more performance or specialized hardware.
How much does AWS Lambda Managed Instances cost?
Billing is based on a fee per million requests plus the underlying EC2 instance costs plus a percentage management surcharge, rather than per invocation and GB-second as with standard Lambda. EC2 Savings Plans and Reserved Instances apply to the instance costs, which can lead to savings versus on-demand Lambda pricing for predictable workloads. Current rates are listed on the AWS pricing page.
Which runtimes and limits does AWS Lambda Managed Instances support?
Supported runtimes include Java 21 and later, Python 3.13 and later, Node.js 22 and later, .NET 8 and later, and Rust. Each function can use up to 32 GB of memory and 16 vCPUs. Scaling is asynchronous based on CPU utilization with no cold starts. If traffic more than doubles within about five minutes, you may experience throttling while capacity scales up.
Note: All product information on this page has been compiled with care, but is provided without guarantee and may be outdated or incomplete. Cloud services evolve rapidly — features, pricing, SLAs, and availability change frequently. Authoritative and up-to-date information can only be found on the official product page of AWS (official documentation). This page does not represent an offer by AWS.